Prediction of paroxysmal Atrial Fibrillation: A machine learning based approach using combined feature vector and mixture of expert classification on HRV signal.

Background And Objective: Paroxysmal Atrial Fibrillation (PAF) is one of the most common major cardiac arrhythmia. Unless treated timely, PAF might transform into permanent Atrial Fibrillation leading to a high rate of morbidity and mortality. Therefore, increasing attention has been directed towards prediction of PAF, to enable early detection and prevent further progression of the disease.